1. Field of the Invention
This invention relates to a method of and system for carrying out an image processing such as a processing for enhancing a predetermined frequency component of an image signal. This invention further relates to a computer-readable recording medium loaded with program for causing a computer to perform the image processing in accordance with the method.
2. Description of the Related Art
We have proposed various image processing methods and systems for improving diagnostic performance of a radiation image by carrying out on a radiation image signal representing the radiation image, for instance, a frequency enhancement processing or a dynamic range compression processing by use of an unsharp mask image signal (will be referred to as “unsharp image signal”, hereinbelow). See, for instance, Japanese Unexamined Patent Publication Nos. 55(1980)-163472, 55(1980)-87953, 3(1991)-222577, 10(1998)-75395, and 10(1998)-171983. For example, in the frequency enhancement processing, a predetermined spatial frequency component of an original image signal is enhanced by subtracting an unsharp image signal Sus from the original image signal Sorig, and adding the remainder multiplied by a coefficient of enhancement β to the original image signal Sorig. This is represented by the following formula (1).Sproc=Sorg+β×(Sorg−Sus)  (1)wherein Sproc is a frequency-enhanced image signal, Sorg is an original image signal, Sus is an unsharp image signal and β is a coefficient of enhancement.
Further, in Japanese Unexamined Patent Publication No. 10(1998)-75395, there is disclosed a method of preventing generation of an artifact in the frequency-enhanced image signal by adjusting the frequency response characteristic of the add signal to be added to the original image signal. In this method, a plurality of unsharp image signals, which are different from each other in frequency response characteristic, that is, in sharpness, are prepared, differences between two of the original image signal and the unsharp image signals are taken, thereby making a plurality of band-limited signals respectively representing frequency components in limited frequency bands of the original image signal, the band-limited signals thus obtained are transformed into signals of desired values by use of different transformation functions, and the add signal is made by adding up the suppressed band-limited signals. This is represented, for instance, by the following formulae (2).Sproc=Sorg+β(Sorg)×Fusm(Sorg, Sus1, Sus2, . . . SusN)Fusm(Sorg, Sus1, Sus2, . . . SusN)=f1(Sorg−Sus1)+f2(Sus1−Sus2)+ . . . +fk(Susk−1−Susk)+ . . . +fN(SusN−1−SusN)  (2)wherein Sproc is a processed image signal, Sorg is an original image signal, Susk (k=1 to N) is an unsharp image signal, fk(k=1 to N) is a transformation function, and β(Sorg) is a coefficient of enhancement determined on the basis of the original image signal.
Further, in Japanese Unexamined Patent Publication No. 10(1998)-171983, there is disclosed a method of preventing generation of an artifact in the processed signal when both the frequency enhancement processing and the dynamic range compression processing are to be carried out. In this method, a plurality of band-limited signals are made in the manner described above, a high frequency component signal representing high frequency components of the original image signal and a low frequency component signal representing low frequency components of the original image signal are obtained on the basis of the band-limited signals, and the frequency enhancement processing and the dynamic range compression processing are carried out by adding the high frequency component signal and the low frequency component signal to the original image signal. This is represented, for instance, by the following formulae (3).Sproc=Sorg+β(Sorg)×Fusm(Sorg, Sus1, Sus2, . . . SusN)+D{Sorg−Fdrc(Sorg, Sus1, Sus2, . . . SusN)}Fusm(Sorg, Sus1, Sus2, . . . SusN)={fu1(Sorg−Sus1)+fu2(Sus1−Sus2)+ . . . +fuk(Susk−1−Susk)+ . . . +fuN(SusN−1−SusN)Fdrc(Sorg, Sus1, Sus2, . . . SusN)={fd1(Sorg−Sus1)+fd2(Sus1−Sus2)+ . . . +fdk(Susk−1−Susk)+ . . . +fdN(SusN−1−SusN)  (3)wherein Sproc is a processed image signal, Sorg is an original image signal, Susk (k=1 to N) are unsharp image signals, fuk(k=1 to N) are transformation functions for obtaining the high frequency component signal, fdk(k=1 to N) is a transformation function for obtaining the low frequency component signal, β(Sorg) is a coefficient of enhancement determined on the basis of the original image signal, and D{Sorg−Fdrc(Sorg, Sus1, Sus2, . . . SusN)} is a coefficient of dynamic range compression determined on the basis of the low frequency component signal, D being a function for transforming D{Sorg−Fdrc(Sorg, Sus1, Sus2, . . . SusN)}.
In the frequency enhancement processing and the dynamic range compression processing (will be representatively referred to “as the transformation processing”, hereinbelow), the frequency response characteristic of the add signal to be added to the original image signal can be adjusted by changing the definition of the transformation functions and the like for transforming the band-limited signals. Accordingly, a processed image signal having a desired frequency response characteristic, e.g., suitable for preventing generation of an artifact, can be obtained by properly defining the transformation functions. However, it is not easy to know how to define the transformation functions in order to obtain a desired result. Therefore, there has been proposed, in Japanese Unexamined Patent Publication No. 10(1998)-63838, a method in which a processed image signal having a desired frequency response characteristic is easily obtained by designating a desired frequency response characteristic for a processed image signal and determining parameters for defining the transformation functions (this parameter will be referred to as “transformation function defining parameter”, hereinbelow) on the basis of the designated frequency response characteristic.
The unsharp image signals used in the aforesaid transformation processing are made by thinning picture elements by filtering picture elements of the original image signal at predetermined intervals and interpolating like number of picture elements. As the filtering processing, a processing for removing high frequency components from the original image signal by use of a low-pass filter, more specifically a processing for calculating an average or a weighted average of the values of picture elements in the filter, has been carried out. In the filtering processing carried out in order to obtain a plurality of unsharp image signals in Japanese Unexamined Patent Publication No. 10(1998)-75395 or the like, the unsharp image signals are obtained by filtering the original image signal to obtain an image signal with less picture elements, further filtering the image signal with less picture elements, and interpolating picture elements into the image signal with less picture elements obtained at each filtering stage so that the number of the picture elements in the image signal becomes equal to that in the original image signal.
Each unsharp image signal is thus made on the basis of the original image signal, which is obtained by reading an original image at a predetermined read density by use of an image reader and digitizing the image signal thus obtained into a digital image signal which can reproduce an image at a predetermined picture element density. It has been known that frequency components lower than a certain frequency determined according to the picture element density (a Nyquist rate) can be correctly reproduced when a digitized image signal is to be reproduced as a printed output. That is, since being determined taking into account the level of image quality required upon reproduction, the read density, i.e., the picture element density is not constant.
For example, in a radiation image read-out and reproducing system, where a radiation image of a human body recorded on a stimulable phosphor sheet is read out as a digital image by scanning the stimulable phosphor sheet with a laser beam, the read density or the picture element density varies depending on the size of the stimulable phosphor sheet and can be freely set by an user.
When image signals different in picture element density or Nyquist frequency are subjected to the same filtering processing using the same low-pass filters and then to the same interpolation, the frequency characteristics of the obtained band-limited signals (more specifically the frequency bands of the obtained band-limited signals) differ depending on the picture element density. This means that, for instance, when a pair of image signals having different picture element densities are obtained by reading an original image at different read densities, and a frequency enhancement processing or a dynamic range compression processing is carried out on the image signals by use of band-limited signals obtained on the basis of the same unsharp image signals, the enhanced frequency band or the compressed frequency band differ between the two original image signals.
In order to overcome this problem, there has been proposed, in Japanese Unexamined Patent Publication No. 10(1998)-63837, a method in which unsharp image signal are obtained by selecting coefficients of filter from a list of coefficients of filter according to information on the picture element density of the original image signal and filtering the original image signal by use of filters of the selected coefficients of filter. That is, when original image signals, for instance, respectively read at read densities of 5 lines/mm and 6.7 limes/mm are filtered by use of the same low-pass filter, the obtained two band-limited signals differ from each other in frequency band. However in the method proposed by the above identified Japanese Unexamined Patent Publication, the two original image signals are filtered by different low-pass filters and accordingly the obtained two band-limited signals can be substantially the same in frequency band. Accordingly, unsharp image signals of the same frequency bands can be obtained irrespective of the picture element density of the original image signals, whereby band-limited signals of the same frequency characteristics can be made and a desired transformation processing, e.g., the aforesaid frequency enhancement processing, can be constantly carried out in the same manner.
However, since energy of a band-limited image represented by a band-limited signal, that is, the peak value of response of a band-limited signal, varies depending on the picture element density, the method disclosed in Japanese Unexamined Patent Publication No. 10(1998)-63837 cannot make constant the response of the band-limited signals in the same frequency band irrespective of the picture element density though can make the frequency bands of the band-limited signals equal to each other. Accordingly, even if a processing is carried out to enhance a band-limited signal in a particular frequency band, the response characteristic of the enhanced band-limited signal delicately varies depending on the picture element density.
In the image processing system described above, there is sometimes input an object original image signal (an original image signal to be processed) representing an image which differs in resolution from images which are normally processed by the image processing system. (The resolution of the images which are normally processed by the image processing system will be referred to as “the reference resolution”, hereinbelow) In such a case, when the transformation processing such as the frequency enhancement processing is carried out on the object original image signal by use of the transformation functions which have been determined for original image signals representing images at the reference resolution, there is a fear that the frequency response characteristic of the image reproduced on the basis of the processed image signal obtained from the object original image signal becomes different from that of the image reproduced on the basis of the original image signal representing an image at the reference resolution. This problem may be overcome, for instance, by preparing a plurality of groups of transformation functions and using one group of the transformation functions according to the resolution of the image represented by the object original image signal. However, this approach is disadvantageous in that the number of the transformation functions becomes too large and management of the transformation functions becomes too troublesome.
Further, though there have been known, as formats for compressing an original image signal, various formats such as JPEG, GIF, TIFF and the like, there is recently proposed a format in which an original image signal is hierarchically decomposed by resolution into hierarchical data, and the hierarchical data in each hierarchy is encoded and compressed. In this compression format, specifically, an original image signal is decomposed by wavelet transformation or the like into a plurality of hierarchical image signals, each having a resolution of ½n times that of the original image signal, and the hierarchical image signals are encoded in the hierarchical sequence and compressed into a single file.
The compression format has the following features.
(1) Since the image signal is not processed block by block unlike the DCT format employed in the conventional JPEG, artifact like block distortion is not generated.
(2) Since the image signals are hierarchically encoded, information on necessary resolutions has only to be transferred upon transfer of the image signals, which results in efficient image transfer.
(3) Since the image signal is decomposed into multiple resolutions, various image processing such as frequency enhancement processing can be relatively easily carried out.
(4) Since space decomposition and frequency decomposition can be simultaneously carried out by multiple resolution analysis, an orthogonal transformation can be carried out over a wide range on a low frequency region, which largely affects the encoding efficiency, whereas over a narrow region on a high frequency region. Accordingly, even if quantization noise is generated around an edge of the image, spatial spread of the noise can be suppressed so that the noise becomes less apt to be recognized.
Further, there have been proposed various file formats such as a FlashPix file proposed by Eastman Kodak in which data of different kinds can be stored in a single file. The aforesaid hierarchical image signals can be also stored in such a FlashPix standard file.
By decomposing an original image signal into multiple resolutions, it is possible to construct the original image signal by a plurality of hierarchical image signals, each having a resolution of ½n times that of the original image signal. This makes it feasible to reconstruct an image on the basis of a part of the hierarchical image signals which is selected according to the image quality required by the output system. That is, when a high quality image is to be reproduced as in a printer, an image signal which can reproduce a high quality image equivalent to the original image in resolution can be obtained by reconstructing the image signal on the basis of hierarchical image signals up to that of the highest resolution. To the contrast, in the case of, for instance, a CRT which cannot reproduce an image in a quality so high as a printer, an image signal which can reproduce an image suitable for the CRT though not so high as the original image in resolution can be obtained by reconstructing the image signal on the basis of hierarchical image signals not including the highest resolution hierarchical image signal and enlarging or contracting the image signal, if necessary.
However since the hierarchical image signals each representing an image lower in resolution (such hierarchical image signals will be referred to as “lower hierarchical image signals” hereinbelow) than that represented by the original image signal differ from the original image signal in frequency response characteristic, if the transformation functions for the original image signal are employed as they are in frequency enhancement processing of the lower hierarchical image signals, there is a fear that an image signal which is different in frequency response characteristic from an image signal obtained by carrying out the frequency enhancement processing on the original image signal can be obtained. This may be overcome by preparing a number of transformation functions conforming to various resolutions and employing transformation functions according to the resolution of the image signal to be processed. However this approach is disadvantageous in that the number of transformation functions to be managed becomes too large and management of transformation functions becomes too troublesome. This problem occurs not only when carrying out frequency enhancement processing on image signals obtained by decomposing an original image signal into multiple resolutions but also when carrying out frequency enhancement processing on an image signal in order to reproduce an image lower than an original image signal.
Further, for example, a radiation image reproduced at pitches of 10 lines/mm by doubling and interpolating an original image signal read out from a stimulable phosphor sheet at a read density of 5 lines/mm is inferior in sharpness to a radiation image reproduced at pitches of 10 lines/mm on the basis of an original image signal read out from a stimulable phosphor sheet at a read density of 10 lines/mm though the sizes of images are the same. This is because high frequency components of the original image signal are weakened depending on the characteristic of the filter employed for changing the picture element density of the original image signal, that is, for contracting and interpolating the original image signal and the characteristic of the filter for doubling the contracted original image signal and because the former image is different from the latter image in frequency response characteristic. Further when a low resolution image is reproduced on the basis of a hierarchical image signal obtained by decomposing an original image signal representing an original image into multiple resolutions, the obtained image becomes inferior to the original image in sharpness depending on the wavelet transformation functions employed in wavelet transformation. Further when a low resolution image is enlarged to the same size as the original image, high frequency components of the original image signal are weakened depending on the characteristic of the filter for enlargement and interpolation and the obtained image becomes inferior to the original image in sharpness. The same problem occurs also when the original image is to be enlarged or contracted to a desired size. Accordingly when the aforesaid frequency enhancement processing is carried out on an image signal representing an image less sharp than the original image, the obtained image becomes different in impression from an image obtained by carrying out the same frequency enhancement processing on the original image signal.